streamlab06: large scale ecogeomorphic flume - conservancy

of 32

StreamLab06: Overview of experiments, instrumentation, and data collection.

National Center for Earth-surface Dynamics

By:

Jeff Marr, Peter Wilcock, Miki Hondzo, Efi Foufoula-Georgiou, Sara Johnson, Craig Hill, Rebecca Leonardson, Peter Nelson, Jeremy Venditti, Ben O’Connor, Christopher Ellis, James Mullin, Anne

Jefferson, Jeff Clark

Date: November 2010

of 32

Acknowledgements

The funding for the StreamLab06 program was provided by the STC program of the National Science

Foundation via the National Center for Earth-surface Dynamics under the agreement Number EAR- 0120914.

Additional funding was provided by the St. Anthony Falls Laboratory, University of Minnesota.

of 32

TABLE OF CONTENTS

ACKNOWLEDGEMENTS 2 1. EXECUTIVE SUMMARY 4 2. PROJECT MOTIVATION 4 3. SUMMARY OF STREAMLAB06 STUDIES 4

3.1 Research Schedule 5 3.2 Ecogeomorphology Study 5 3.3 Gravel Augmentation and Patch Dynamics Study 6 3.4 Sand Infiltration Study 6 3.5 Aggregation/Degradation Study 7

4. THE MAIN CHANNEL FACILITY 7 5. BED MATERIAL 9 6. INSTRUMENTATION AND DATA COLLECTION 10

6.01 Backbone Data 11 6.02 Sediment Flux Monitoring System 12 6.03 Sediment Flux Data 13 6.04 Flow Field 13 6.05 Topography 14 6.05.1 Flume Coordinate System 14 6.05.2 FlowOn Toposcans 14 6.05.3 FlowOff Toposcans 14 6.06 Color Photography 15 6.07 NIR Photography 16 6.08 Bed Facies Mapping 16 6.09 Grain Size Characterization 17 6.10 Passive Radio Frequency Identification (RFID) 18 6.11 Hydraulic Conductivity 18 6.12 Conductivity Probes 18 6.13 Water chemistry/characterization 18 6.14 Subsurface water temperature 18 6.15 Nutrient uptake 19 6.16 Ecological measurements 19

7. DATA ARCHIVAL AND ACCESS 20 8. CONCLUSIONS 21 APPENDIX I: REFERENCES 22 APPENDIX II: RESEARCH TEAM MEMBERS 23 APPENDIX III: DEFINITIONS 24 APPENDIX IV: MAIN CHANNEL SAFETY PLAN 25 APPENDIX V: DATA FILE FORMATS 26

FlowOn & FlowOff Toposcan Data 26 ADV Processed file. Used WinADV software to process .ADV files (raw data) 27 Flourometer output data 28 Hydrolab Output Datafile 29 Autoanalyzer Output Data 30

APPENDIX VI: SEDIMENT FLUX POST PROCESSING 31 Sediment flux Data Archival Organization 31 Converting Raw Data into Input Data: Pre-processing steps 31 MatLab Scripts 32 Final Data 32

of 32

1. EXECUTIVE SUMMARY

This report summarizes the StreamLab06 experimental research program conducted in the St. Anthony Falls

Laboratory (SAFL) Main Channel facility from April through October 2006. The experiments were funded

through the National Center for Earth-surface Dynamics and involved a host of researchers, graduate students,

visitors, and undergraduate students. The experiments were organized into seven phases of work. The first two

phases of the project involved testing of conventional and surrogate bedload monitoring technologies (Marr et. al.

2007). The last five phases involved interdisciplinary research of sediment transport and ecohydraulics. This

report focuses on the later phases of the project and does not include the bedload monitoring technologies.

This report contains information on the organization of the experiments, the methodologies and protocols used

to collect data, the types of data collected, data structure and format, and information on data storage and access.

2. PROJECT MOTIVATION

Experimental studies of river processes must negotiate a tradeoff: in order to gain control of essential variables

and support tractable technical measurements, experiments are typically conducted at a reduced scale, at the cost

of some loss of realism. The methods and limitations of scale modeling are well-developed in some aspects of

river science, such as hydraulics (ASME, 1929) and sediment transport (Sheng, 1987), and are emerging in others,

such as biochemical transport and flow/organism interactions (Arnon et al., 2007, Hondzo and Wang, 2002). Some

features of natural systems are difficult or impossible to scale. These include processes involving aquatic

organisms and their interactions with their physical surroundings, as well as features, such as channel-scale bed

forms and habitat, that arise from interactions among processes operating across a range of spatial scales. A sound

understanding of both local mechanisms and broader interactions is needed to develop predictive models in river

science (Paola et al., 2006). The solution is to conduct experiments at full scale while maintaining experimental

control and using instrumentation that can resolve both local and full-scale processes. Such an approach imposes

considerable technical and conceptual challenges.

We have initiated an experimental program, StreamLab, of full-scale experiments on linked

physical/chemical/biological processes. StreamLab has been developed at the National Center for Earth-surface

Dynamics (NCED) and the St. Anthony Falls Laboratory (SAFL). The program’s essential features are an explicit

multi-disciplinary focus, experimental control at the field scale, and the use of advanced technology to support

detailed observations typical of small-scale lab experiments. The work is motivated in part by the need to provide

better science to support the large public investment in stream rehabilitation and restoration. Hence, an initial

focus of the work is the ecosystem response to physical alterations to the stream system.

The StreamLab06 experiments utilized the Main Channel facility at the St. Anthony Falls Laboratory.

Extensive upgrades to the Main Channel’s recirculation and monitoring system in 2005 paved the way for this

interdisciplinary project. The overall scope of work undertaken in StreamLab06 is large and served to test the

boundaries of the facilities and the experimental conditions. The data from StreamLab06 were collected within

four well-defined research studies, which are summarized in the following section. Many of the individual datasets

may be of use to other researchers, and are available to the public through the NCED Data Repository

(www.repository.nced.umn.edu). The repository allows viewing and/or download of any of the available datasets,

which include:

Hydraulic conditions (discharge, water slope, bed slope, depth average velocity, and flow field

mapping)

Physical conditions (bed topography, bar locations and shape, photo imaging of bed)

Bed characterization (surface and subsurface grain size distribution (GSD), patch location and GSD,

surface patch topography and images)

Sediment transport characterization (continuous sediment flux, recirculation GSD)

Water chemistry (temperature, dissolved oxygen, pH)

Biological conditions (heterotrophic respiration, biomass accumulation, nutrient processing rates)

3. SUMMARY OF STREAMLAB06 STUDIES

There were four studies conducted within the Main Channel as part of the StreamLab06 project (Table 3.1).

The studies all shared the same sediment and general experimental configuration. The largest study,

http://www.repository.nced.umn.edu/

of 32

Ecogeomorphology Response to High Flow Events, involved examination of the dynamic response of various

physical, biological, and chemical parameters to large and small flood perturbations. This phase will henceforth

be referred to as the Ecogeomorphology Study. The Armoring and Gravel Augmentation Study focused on the

linkages between sediment supply on reach-scale morphodynamics, in particular surface patch and bar formation

and maintenance. The study also examined the river restoration application of gravel augmentation designed to

re-establish sediment supply to a sediment-limited reach. The Sand Infiltration Study examined the processes of

sand wave propagation, infiltration, and co-deposition in the coarse bed material within an alternate bar

morphology and heterogeneous flow field. Finally, the Aggradation/Degradation Study examined the sediment

transport and hydraulic process of major channel grade adjustment under falling base level (degradation) and

rising base level (aggradation).

A fifth study was conducted but not reported here. Originally called Phase 1 and Phase 2 of StreamLab06, this

study involved the ground-truth testing of various conventional and surrogate bedload monitoring technologies.

The work was done in collaboration between NCED, SAFL and various federal agencies and consultants. The co-

organizing agency was the US Geological Survey (John Gray). Phase 1 of this work involved a sand bed channel

and Phase 2 involved a clean gravel material. The backbone data system (described in Section 6) was in operation

during these runs, which were performed in January-March 2006. Sediment flux data, water temperature,

discharge, flow depth, and limited flow velocity information is available from these runs which involved five

discharges with the sand bed and four discharges with the gravel bed.

3.1 Research Schedule The order and schedule of the StreamLab06 studies was complex and required the research team to strictly

adhere to a schedule. The last component of the Ecogeomorphology Study involved growing periphyton alga on

the bed of the channel and observing the influence of the biofilm on various physical and chemical stream

characteristics. Because the growth of the periphyton is sensitive to water temperature, it was imperative that this

phase of the experiment was completed before the river water temperatures dropped too low. The bed material

grain size distribution dictated the order of the experiments. It is physically easier and cheaper to add sand to a

bed versus removing sand. This meant that the clean gravel runs needed to occur prior to the sandy gravel runs.

Also, gravel augmentation and sand infiltration runs needed to occur in the clean gravel bed material.

Table 3.1. Overview of the four StreamLab06 studies detailed in this report

3.2 Ecogeomorphology Study The goal of the Ecogeomorphology Study was to evaluate the effect of topographic and grain-size complexity

on sediment and solute transport, and their interaction with periphyton biomass and spatial distribution. Channel-

scale features such as bars and pools introduce spatial variability in topography and bed composition, which in

turn controls local flow, transport, and near-bed ecohydraulic processes. This study included experiments with a

plane-bed configuration as well as more complex alternate bar topography with local sediment sorting. High-

resolution observations of bed topography and grain size, flow, and transport were combined with integral

measures of water biochemistry to evaluate the effect of channel complexity on sediment transport and sorting,

Phase Bed Material Morphology no vegetation

Ecogeomorphology Response to High Flow Events

Phase 3a clean gravel plane bed no periphyton

Phase 3b clean gravel alternate bar no periphyton

Phase 5a sandy gravel plane bed no periphyton

Phase 5b sandy gravel alternate bar no periphyton

Phase 6 sandy gravel alternate bar periphyton

Gravel Augmentation and Patch Dynamics Study

Phase 4a,b clean gravel alternate bar Not Applicable

Sand Infiltration

Phase 4c clean gravel alternate bar Not Applicable

Aggradation/Degradation

Phase 7 sandy gravel plane bed Not Applicable

of 32

hyporheic exchange, nutrient retention, and community respiration. Two periods of periphyton growth were

included to support observations of the interaction among bed configuration, sediment composition, heterotrophic

biomass accumulation, hyporheic exchange, nutrient retention, and dissolved oxygen profiles near the sediment-

water interface. These experiments thus provided a template to examine the effect of bed composition (sandy or

clean gravel), bed topography (plane or alternate bar), and transport rate (moderate or high), on surface and

subsurface grain sorting, surface and subsurface water storage and flow paths, autotrophic and heterotrophic

biomass accumulation, metabolic rates, and the uptake and retention of ecologically important nutrients.

3.3 Gravel Augmentation and Patch Dynamics Study Gravel augmentation, the addition of gravel into a stream, is a common remobilization strategy used in alluvial

river systems. Typically, gravel is added (augmented) into rivers downstream of dams, where natural sediment

supplies have been reduced. A reduction in sediment supply has negative physical, biochemical, and ecological

consequences, such as a reduction in available habitat for fish spawning and rearing (Kondolf, 1997). It can also

lead to the reduction of hyporheic exchange with the subsurface, potentially leading to changes in water

temperature, nutrient flowpaths, and bioavailability in the near-subsurface (Poole and Berman, 2001). The Gravel

Augmentation and Patch Dynamics Study investigated the use of a fine-grained augmentation as opposed to a

coarser spawning size gravel augmentation to essentially reset the river back into a mobile bed regime. Our

approach was based on the fact that under the static armor layer there is an abundant source of sediment with sizes

that span the entire distribution of the river system. We hypothesized that a fine grained augmentation would serve

to “smooth” the coarse bed by filling in spaces between the immobile armor, reducing the roughness of the bed.

We were interested in determining if this would increase near-bed shear stress velocities and drag forces on large,

immobile particles to a level that would break up the surface armor layer and provide access to finer grains in the

subsurface, reducing the amount of coarse gravel augmentation required.

In conjunction with the gravel augmentation experiments, we investigated the development of bed surface

grain size patchiness. Patches are a phenomenon associated with gravel-bed rivers and understanding their

dynamics is critical to stream restoration for many reasons, one of which is the fact that the river bed is the

microenvironment in which the stream food web begins. In StreamLab06 we focused on forced patches: persistent

stationary sorting features associated with bed topography. The Main Channel facility enabled us to develop self-

formed bed topography with large sediment sizes and flow depths, such that detailed hydraulic measurements

were possible. The primary goal of this project was to examine where patches formed in the flume relative to bar

position and relative to measurements of divergent boundary shear stress resulting from the bars. The data

collected will be used to further test hypotheses and numerical models on the effect of shear stress divergences on

surface sorting.

We note that this study was an extension to ongoing experiments conducted as part of the California Bay-Delta

Authority project Physical Experiments to Guide River Restoration and highlight the collaborative nature of the

StreamLab concept. Through careful planning and experimental design, StreamLab06 was able to provide an

additional set of experiments to the CalFed project. The data and results from these experiments will be used to

improve the utilization of gravel augmentation in stream restoration.

3.4 Sand Infiltration Study The transport, infiltration, and co-deposition of fine-grained sediments (sand, silt, and clay) in gravel-bed rivers

are also important topics for stream restoration. These processes can lead to ecohydraulic problems, especially in

areas where there is an excessive supply of fines (such as downstream of a dam removal site), in regions of recent

logging activity or forest fires, or in watersheds with active mass wasting processes (e.g. landslides or debris

flows). The Sand Infiltration Study involved a set of experiments studying the relationships between bed

topography and subsurface fines. The first experiment was a survey of fines deposited with a high gravel transport

rate as alternate gravel bars were being formed. There appeared to be a correlation between some topographic

features, such as the lateral edge of gravel bars, with higher subsurface fines content. In the second experiment,

fine sand was infiltrated into a gravel bed with alternate bar topography to test whether infiltration relationships

determined in 1-D (plane-bed) and 2-D (dune) flume experiments predicted infiltration into a bed with 3-D

topography. Experiments involved detailed pre-infiltration measurements of sand concentration followed by

infiltration of fine sand and post-infiltration measurements. Initial results from the work support past findings on

infiltration (e.g. Diplas and Parker, 1992; Lisle, 1989; Elliott and Brooks, 1997). Infiltration of fine sand resulted

in a sand seal; a high concentration layer of sand that forms near the surface, thereby stopping further penetration

of 32

of fines. Infiltrated fines reached 8-10 cm below the bed surface, which is at the deep end of expected values (2-

3*D90, where D90 = 14mm). There was moderate spatial variability of infiltration depth that appeared to be

correlated to topography. Sand infiltration was greatest in the deepest parts of the channel as compared to the sides

of the channel (bar tops). Subsurface flow in both the lateral and streamwise directions appeared to be affecting

infiltration subsurface sand patterns. One example is sand found underneath emergent (dry) gravel at the

downstream end of a bar. The infiltrated sand in this location suggests that sand traveled laterally in the subsurface

by lateral hyporheic flow pathways. The data collected in this phase will be used to further develop theories and

numerical models of sand infiltration in heterogeneous flows.

3.5 Aggregation/Degradation Study Rivers often encounter major perturbations that result in significant re-grading of the channel bottom.

Examples of perturbation include input of sediment slug or pulses, major changes in base level, or sudden changes

in water discharge. The Aggredataion/Degredation Study utilized the Main Channel’s sediment recirculation

system and tail-water control to explore the processes of major aggradation and degradation in a channel. Two

runs were conducted. A wealth of topographic, sedimentological, and hydraulic data was collected. The first run

involved hydraulic degradation (erosion) of the upper half of the flume test section and deltaic deposition in the

downstream half of the flume. The second run involved hydraulically eroding the downstream portion of the

channel and, via the recirculation system, progradational deposition at the upstream half of the channel. Careful

measurement using RFID tagged rocks, surface sampling for grain size, and continuous topographic scanning

were performed during this run.

4. THE MAIN CHANNEL FACILITY

The StreamLab06 studies were conducted in the Main Channel facility at the St. Anthony Falls Laboratory

(Figure 4.1). The channel has a rectangular cross-section that measures 2.74 meters in width and 1.8 meters in

depth. Water for the channel is sourced from the Mississippi River by controlled diversion of water through

SAFL’s intake structure. The maximum discharge in the channel is 8.5 m3/s. Approximately 55 meters

downstream from the entrance of the channel is the Sediment Monitoring and Recirculation System (SMRS – see

Section 6) and 15 meters downstream of the SMRS is a sharp crested weir with the dual purpose of controlling

tail water elevation and instrumented monitoring of water discharge. A schematic SAFL’s Main Channel Facility

is shown in Figure 4.1.

Figure 4.1. Section schematic of SAFL Main Channel.

The SAFL Main Channel has the ability to recirculate large quantities and large sizes of sediment 55 meters

upstream of the SMRS; allowing long duration sediment transport research. The recirculation system is capable

of moving particles up to 75mm (3 in) in diameter. The system was originally designed in the early 1980s by the

Federal Interagency Sedimentation Project as part of a program for laboratory testing of the 3” Helley Smith

Sampler (Hubbell et al, 1987). The intake for this system is located in the same floor-pit used to accommodate the

VERTICALENTRANCE

TUNNEL

WATER INTAKE (SLUICE GATE)

SHARPCRESTED

WEIR

SEDIMENTMONITORING &

RECIRCULATIONSYSTEM

P

SEDIMENT RECIRCULATION PIPE

TEST SECTION ( 55 m )

15.2 m

65 m

FIGURE 4.2 LIMITS

KNEE WALL

of 32

weigh drum system (see Section 6.02). At the bottom of this pit is a large auger shaft that is attached to a variable

speed hydraulic motor. Rotation of the auger causes gravel to move laterally in the pit toward an outlet located at

the river right side of the flume, and into the recirculation pump intake. The recirculation of sediment is

accomplished by a large 3-phase recessed impeller centrifugal pump. The pump has an 8-inch intake and outlet

and requires approximately 0.25 m3/s of water. It is important to note that the water used by the pump to transport

material upstream is not supplied from the research channel. If water was withdrawn from the research channel it

would induce a downward flow of water that would 1) potentially result in the suction of sediment into the weigh

drums, and 2) add a false loading onto the weigh drums. Water for the pump is therefore supplied from a separate

source. A schematic of the recirculation system is shown in Figure 4.2.

Figure 4.2. Section detail of the Sediment flux and recirculation system

A new installation on the Main Channel was the design, fabrication, and installation of a three-axis positionable

data acquisition (DAQ) carriage (Figure 4.3). The DAQ carriage is capable of traversing the entire 55 x 2.74 meter

test section and can position probes to within 1 mm in all three axes. Travel speeds are up to 2 meters/second. The

DAQ carriage is controlled by a “backbone” computer that also serves as the master time clock for all data

collection in the study. The DAQ carriage is used to position a number of data collection technologies such as

topographic and bathymetric surveys, photography, and acoustic Doppler velocimetry (ADV). Complete details

on all the instrumentation used to collect data in StreamLab06 are provided in Section 6.

of 32

Figure 4.3. Three-axis automated data collection carriage

5. BED MATERIAL

The bed material for the StreamLab06 experiments was donated to St. Anthony Falls Laboratory by Aggregate

Industries North Central Region Division. The channel’s base mix, referred to as “clean gravel,” is a custom

distribution that consists of three readily available local aggregates that were mixed at a ready-mix facility in a

specified ratio. This distribution was created by testing combinations of available aggregates in various ratios until

a distribution was created that satisfied a unimodal lognormal distribution with a mean of 8mm and a standard

deviation of 2mm. The final distribution for clean and sandy gravel mixtures is shown in Figure 5.1.

Figure 5.1. Grains size distribution for the clean gravel and sandy gravel bed material.

The custom mix (clean gravel) was delivered to SAFL in eight truckloads of about 24 tons each. The sediment

was then hydraulically loaded into the Main Channel facility via an 8 inch pipe traversing two levels of the

laboratory. The sandy gravel bed was created by adding 15 tons of sand to the existing gravel mixture and both

StreamLab06, Bed Material Grainsize Distributions

0

10

20

30

40

50

60

70

80

90

100

0.1 1.0 10.0 100.0

diameter (mm)

Cu

mm

ula

tiv

e P

erc

en

t F

ine

r

Clean Gravel

Sandy Gravel

of 32

hydraulically and manually turning the bed over. Trenches were excavated and refilled using shovels and a small

bobcat until the flume was mixed. The distribution of the sand gravel mix is shown in Figure 5.1. Surface and

subsurface bed material samples were collected throughout the research.

6. INSTRUMENTATION AND DATA COLLECTION

The StreamLab06 experiments utilized an array of advanced technologies to monitor the physical, biological

and chemical conditions in the channel. In this section we provide technical detail of the various technologies used

throughout the various studies. This section does not provide specifics on the deployment protocols of an

instrument within a research study; rather we focus here on the general characteristics of an instrument (i.e.

sampling rates, accuracies, data formats). A summary of all the technologies discussed here is given in Table 6.1.

All data collected in the StreamLab06 experiments is available to the public through the NCED Data Repository

(www.repository.nced.umn.edu).


of 32

Table 6.1. Summary of instrumentation and data acquisition system used in StreamLab06.

6.01 Backbone Data A variety of data was collected during the entire StreamLab06 experiments and this data is referred to as the

Backbone Data. The data was logged by the Central Data Acquisition Computer (CDAQ), which served as the

master time clock for the experiments. This master time was adopted by the DAQ carriage computer system as

well. The CDAQ was programmed to log several pieces of data that included water temperatures, water discharge,

sediment accumulation in the five weigh-drums, and downstream gate elevation. The data are sampled by the

CDAQ system at 5000 Hz for 0.8 seconds. These values are then averaged to a single value, which is written to

ASCII data file at 1 Hz. In other words, for each of the data systems wired to the CDAQ system (weigh drums,

water temperature, tail water elevation, sharp-crested weir elevation) an averaged data point is written to a master

data output file at ~1 Hz. These data are archived as part of the StreamLab06 backbone dataset. More information

on these files is given below.

Water Temperature: Water temperature was monitoring using a YSI thermistor capable of measuring to +/-

0.1 degree C.

TECHNOLOGY DESCRIPTION

Backbone Data

water temperature continuous water temperature record

water discharge continuous water discharge record

sediment flux continuous total bed material flux out of test sections

downtream tailwater elevation continuous record of tailwater elevation in channel

Flow field

acoustic Doppler Velocimeter vertical profiles and cross-sections; turbulence measurments

micropropellers near-bed velocity profiles under low flow conditions

surface flow markers (confetti) velocity estimates at shallow flow conditions

Topography

water surface profiles water surface topography measured at low and high flows

bathymetric profiles bed surface topography measured at low and high flows

bed digital elevation mapping detailed bed-surface maps collected post-flood

patch topography detailed topographic laser scans of identified surface patches

Photography

digital surface photos complete digital photo coverage of bed surface post-flood

Bed facies and Grain Size Analysis

facies maps hand-drawn facies maps of bed surface post-flood

Grain Size Analysis

bulk bed samples bulk sampling and sieving of bed material

gravel patch characterization Klingeman grain size analysis of surface and subsurface at identified patches

recirculation samples collected samples at random intevals at discharge point of sediment recirculation

Water Quality Characterization

general water quality data near-continuous record of water temp, dissolved oxygen, conductivity and PAR

specific water quality data periodic grab samples analyzed for nitrogen and phosphorus

Gravel Transport Dynamics

Dynamics tracer studies (RFID) stationary and mobile Radio Frequency Identification systems for tracking rock movement

Static tracer studies (Colored Stones) install and retrieval of painted stones to track surface transport

Subsurface Characterization

Permeability comprehensive post-flood measurement of permeability post-flood

Conductivity probe nests subsurface flow path mapping using salt injections and conductivity sensor arrays

subsurface auto-logging thermistors four vertical arrays of autologging thermistors inserted into alternate bar bed

Surface/subsurface Interaction

tracer injections and monitoring Salt and dye injections used to determine residence time of subsurface flow

Biomass Characterization

Surface and subsurface scrape samples Surface and subsurface sampling of bed material for biomass

NIR photography with calibration Beta testing of near Infrared camera to detect biomass growth in flume

of 32

Water Discharge: The elevation of the downstream sharp crested weir, Hweir was measured with a linear

distance sensor. The downstream water surface (tailwater) elevation, Hwater was measured with Massa sonic

range finder. Water discharge was determined from a calibrated sharp-crested weir equation (Equation 6.1).

(6.1)

Where,

w = channel width of 2.74 m

K = 1838, sharp-crested weir coefficient for units of liters per second

Hwater and Hweir are measured in millimeters

Sediment flux: The Backbone Data also includes record of sediment accumulation in the sediment flux

monitoring system. Technical details on this system are given below. Accumulated weight in the five weigh-

drums was recorded along with the other backbone data and written to a single text data file with time and

date stamps for the experiments.

6.02 Sediment Flux Monitoring System An essential feature of the Main Channel is a sediment flux monitoring system that was designed, fabricated,

and installed by SAFL in 2005. This system is capable of continuously monitoring the particle flux (bedload)

during an experimental run. The monitoring system is made up of five weigh drums that collect, weigh, and

periodically empty (rotate) transported sediment; each drum collects gravel from 0.55m of channel width. The

drums are positioned side-by-side under the floor of the channel in a pit. Removable stainless steel cover plates

with 3” slots are installed over the drums and serve to funnel the gravel directly into the drums. The drums are

constructed of aluminum and have three radial baffles welded at 120 degrees to each other. The rotation axis of

the drum is aligned parallel to the water surface and transverse to the flow. Each drum is supported by two

aluminum arms that extend vertically upward toward the ceiling. This structure is then attached, by chains, to a

single load cell. The system uses five load cells manufactured by Interface Advanced Force Measurement (SM-

250) and has a maximum load of 250 lb and an accuracy of 0.121 lb. In order to not exceed the maximum capacity

of the load cell, the weigh drums are designed to empty when the accumulated weight exceeds a user-specified

weight (20-40kg). Rotation of the drum is driven by a pneumatic piston that, when extended, drives the drum 120

degrees clockwise and, when retracted, 120 degrees counterclockwise. Each of the load cells is monitored by a

central data acquisition system (CDAQ). The CDAQ system monitors the five weigh pans, water temperature, tail

water elevation, and sharp crested weir elevation, and writes these data to a single output file for the experiment.

The data are sampled by the CDAQ system at 5000 Hz for 0.8 seconds. These values are averaged to a single

value, which is written to ASCII data file at 1 Hz. In other words, for each of the data systems wired to the CDAQ

System (weigh drums, water temperature, tail water elevation, sharp-crested weir elevation), an averaged data

point is written to a master data output file at ~1 Hz.

The weigh drums are used to estimate flux by recording the time rate of change in sediment accumulation in a

single weigh drum. As sediment is transported out of the test section and is trapped in the weigh drums, the load

cells record the ever-increasing weight of material in the drums. Periodically, the weigh drums reach their

maximum load and a tipping event is initiated, and the accumulation process starts over. Post processing of the

data, including removal of the tipping events, allows estimation of sediment flux at each of the five pan locations.

5.2

1000**

weirwater

w

HHwKQ

of 32

Figure 6.1. Section schematic of weigh drum

6.03 Sediment Flux Data Data collected from the sediment weigh-drum system is recorded, as described above, in the backbone

dataset in a file with the following naming scheme:

Sediment Flux Data: aaaalps_yyyymmdd_sedflux_[bedmaterial]_[morphology].csv

where, aaaa is the design discharge in liters per second, [bedmaterial] is either cleangravel or sandygravel, and

[morphology] is either plane bed or alternate bar. The raw sediment flux data contains five columns of

accumulated gravel in the weigh-drums (drums 1 – 5) as well as five columns of real-time sediment flux estimates

computed during the data collection. The raw data also contains the weigh-drum dumping events that occur from

time to time.

The sediment flux files are a key dataset because they contain the official time stamp, discharge history,

sediment flux rates, and hydraulic stage information for the runs. This data can be used to accurately determine

when a run started and stopped, as well as potentially explain experimental anomalies. This dataset is the most

important in terms of coordinating with the other various datasets. For example, ADV measurements and

topography information can be linked together with the sediment flux data through the official times recorded in

the sediment flux files. The archive of sediment flux data is viewed to be the most important of all the data since

it contains the official times of the run.

The sediment flux data files have been reviewed and post-processed to a final form, and are available in several

formats. A complete description of the sediment flux data and post processing is available in Appendix V.

6.04 Flow Field Detailed surveys of the velocity/flow field and turbulence were made in the flume during some of the runs.

Measurements were taken with acoustic Doppler velocimeters (ADV) and Shinozuka micropropellers. Both 16

MHz and 10 MHz Sontek ADVs were used. The data includes both vertical profiles and cross-sectional details

for both plane bed and alternate bar topography.

of 32

6.05 Topography Several methods were used to record topography, bed slopes, and water slopes during the StreamLab06

experiments. We divide the acquisition into “FlowOn” and “FlowOff”, where FlowOn indicates that water was

flowing in the channel and FlowOff indicates a dry bed. With the exception of some hand survey with level rods,

most of the topographic and bathymetric surveys were done from the Data Acquisition Carriage (DAQ Carriage

– see Section 4). The DAQ carriage surveys were performed using a series of user-defined scripts. Scripts are

programmed instructions that direct the DAQ Carriage’s movements, speeds, and data collection protocols. For

the most part, we used a standard set of scripts for all research in StreamLab06, with only slight modifications in

cases of complex bathymetry and shallow flows. All the DAQ scripts are archived in the StreamLab06 data archive

on the NCED Data Repository (www.repository.nced.umn.edu).

6.05.1 Flume Coordinate System The coordinate system was defined for the flume and was the basis for all topographic survey and position of

data in StreamLab06. The origin (0,0,0) of the flume was the far upstream, river left corner of the flume. The x-

direction of the flume is in the streamwise direction with zero at the upstream end of the flume and 55,000

millimeters at the upstream face of the weigh-drum system. The y-direction is lateral to flow with zero along the

river left wall (looking downstream) and 2740 millimeters at the river right wall. The z-direction is the vertical

axis with zero at the top of the stainless carriage rail along the river left wall and positive downward. All distances

in the flume are referenced in millimeters.

6.05.2 FlowOn Toposcans When water was flowing in the channel, we captured both water surface and bed surface surveys. The DAQ

carriage allowed simultaneous capture of the water surface and bed surface profiles. Water surface was measured

using a Massa sonic range finder and the bed surface water was measured using high-frequency submersible bed

sonar. Both devices are accurate to < 1mm. Both the Massa and the submersible sonar were mounted to the mast

arm of the DAQ carriage and data were logged from these devices simultaneously at the spatial grid nodes

specified in the user-defined script. The FlowOn toposcans typically involve five survey lines or “passes” of the

flume and three lateral survey lines (transects located at x = 9300mm, 30000mm, and 54450mm) at a grid spacing

of 10 x 10 mm covering the majority of the test section (9000mm to 52000 mm). The data from the FlowOn survey

was written to two separate comma delimited data files (.csv) with the following file-naming scheme:

FlowOn water surface: aaaalps_yyyymmdd_flowontopo_[bedmaterial](MAS)_scan00bb.csv

FlowOn bed suface: aaaalps_yyyymmdd_flowontopo_[bedmaterial](SNR)_scan00bb.csv

Where aaaa is the design water discharge in liters per second and bb is the scan number, which represents the

number of times the script was executed during the run. In other words, bb represents the number of time the

script has been executed at a certain discharge. For example, if the flow was turned on at the start of the day to a

fixed discharge and the research team ran the FlowOn toposcan every 60 minutes over the 6 hours of run, there

would be filenames with *_scan0001.csv through *_scan 0006.csv. [Bedmaterial] is either cleangravel or

sandygravel (bed material) and planebed or altbars (morphology). Details on the file structure of the FlowOn

toposcans are given in Appendix IV.

6.05.3 FlowOff Toposcans Detailed surveys of the bed surface were recorded using a Keyence range finder, which has a vertical accuracy

of < 1mm. The Massa sonic range finder was also used. Three FlowOff scripts were used frequently:

FlowOfftopo10x10 10 x 10 mm survey of the bed to generate digital elevation maps. Typically

uses both Keyence and Massa devices and output is one *.csv file from

each device (two files total).

FlowOfftopo5pass1mm Five streamwise passes in the flume but with a data point every 1 mm in the

x-direction. Typically uses both Keyence and Massa devices and output is

one file from each device.

FlowOffpatch1x1 Patch scans of the bed done at 1 x 1 mm grid typically over a 300 by 300

mm area of the bed. Boundary of patch is specified in the filename. Scans


of 32

are done with the Keyence laser and output is a comma delimited (CSV)

text file.

Datafiles from the FlowOff toposcans for Keyence and Massa acquisition employ the same naming scheme

described above, where:

FlowOff bed surface: aaaalps_yyyymmdd_[scriptname]_[bedmaterial](KEY)_scan00bb.csv

FlowOff water surface: aaaalps_yyyymmdd_[scriptname]_[bedmaterial]_(MAS)_scan00bb.csv

Where aaaa is the design water discharge in liters per second and bb is the scan number. [scriptname] is the

name of the script: FlowOfftopo10x10, FlowOfftopo5pass1mm, FlowOffpatch1x1. [bedmaterial] is either

cleangravel or sandygravel (bedmaterial) and planebed or altbars (morphology). Details on the file structure of

the FlowOff toposcans are given in Appendix V.

6.06 Color Photography Digital photo documentation of the flume surface was made along with FlowOff digital DEM Scans

(FlowOfftopo10x10). The photos were taken with a Nikon D70 digital SLR camera mounted to the

instrumentation mast on the DAQ Carriage. A script was written to move the camera across the entire test section

surface. A high resolution photograph was captured every 0.5 meters in the streamwise (x) direction. Two photos

were captured at each x position. A post processing application has been programmed by David Olsen using Visual

Basic.net to stitch the photographs together into a single mosaic (Figure 6.2). The photo-stitching application is

available with the StreamLab06 data on the NCED Data Repository. 182 images were captured for each photo-

survey that included the flume from 10.000m to 54.500m. Photos used the following file-naming scheme:

Color Photos: aaaalps_yyyymmdd_flowoffphotos_[bedmaterial](X_#, Y_#, Z_#)Imgbbbb.JPG

Where aaaa is the design water discharge in liters per second and bbbb is the image number (1-182).

[bedmaterial] is either cleangravel or sandygravel.

of 32

Figure 6.2 Example of stitched photoscan involving the 182 images collected during a FlowOff photoscan

6.07 NIR Photography A small subset of photographs were taken using a digital SLR camera configured for capturing near-infrared

images of the bed surface. This effort was a pilot study to determine if it was possible to track biomass growth in

the flume using NIR light wavelengths. The approach is based on the idea that alga and vegetation emit light

frequencies in the range of NIR and, using a properly designed capture technique, it may be possible to broadly

characterize and track growth of biomass in the flume. Photos taken during such surveys follow the file naming

format:

NIR Photos: yyyymmdd_flowoffphotos_IR_[phasestage](X_#, Y_#, Z_#)Imgbbbb.JPG

6.08 Bed Facies Mapping Visual surveys of the bed surface and bed facies were typically made after the flow was turned off. Bed facies

maps were hand-drawn and delineated major grain size patches, bar locations, bedforms, and other observations

worth noting. The facies maps also indicate approximate location of where patch samples and other physical

measurements were taken. Both digital and hard copy bed facies maps exist. Data files for digital bed facies maps

use the following file-naming scheme:

Facies Map: bed facies map_yyyymmdd.pdf

Patches delineated by dominant grain size were labeled using the following nomenclature. Capital letters

indicate highly present grain sizes; lower case letters indicate less present grain sizes. Letter appearing first

indicates more present than the following letters. Examples are provided below.

Cmf: Coarse grains dominate patch. Medium and fine grains present; medium grains more

present than fine grains.

Fmc: Fine grains dominate patch. Medium and coarse grains present; medium grains more

present than coarse grains.

MC: Medium and coarse grains highly present throughout patch. Little to no fine grains present.

An example of a full facies map is shown in Figure 6.3. Examples of correlations between facies maps and

digital photoscans are shown in Figure 6.4.

of 32

Figure 6.3 Example of a bed facies map

Figure 6.4. The facies maps and the digital photographs can be used together as shown in this example

6.09 Grain Size Characterization A variety of grain size samples were collected throughout the StreamLab06 research. Standard wet and dry

sieve procedures were adopted for the analysis. The main categories of samples were:

of 32

Bulk bed samples - bed material collected at the beginning of each phase and sieved by standard

sieve protocols at half-phi intervals.

Recirculation samples – samples collected at the upstream end of the flume at the exit of the

recirculation pipe.

Patch samples – samples collected of the surface and subsurface using a Klingeman sampling

protocol and used to characterize surface patches. Patch sizes were 30 cm x 30 cm. Prior to

removing surface material, a detailed 1mm x 1mm FlowOff surface scan was performed (see

Section 6, FlowOff toposcans).

Biomass surface and subsurface sampling – surface and subsurface sampling of bed material by

ecology group. Primary use of samples was to measure biomass accumulation, but grain size

analyses of these samples were also performed.

6.10 Passive Radio Frequency Identification (RFID) Nearly 600 rocks ranging in size from 20 to 30 mm were washed, core-drilled, painted, and inserted with a

passive Radio Frequency Identification (RFID) tag. The tagged rocks were used in a number of StreamLab studies.

A main advantage of the technology is the passive tags are relatively cheap and do not require batteries. The tags

receive power when they pass through a strong magnetic field induced by coiled wires, i.e. antennas. Once

powered, the tags emit a unique identification, which is recorded along with time by a PDA datalogger.

The StreamLab06 studies employed two configurations for using RFID. The first was stationary antennas that

were buried in the sediment bed and tracking rocks passing over top. The second configuration involved a mobile

antenna attached to the underside of the DAQ carriage. The antenna was powered on and slowly swept over the

test section. Rocks were located and positions recorded.

6.11 Hydraulic Conductivity The hydraulic conductivity of the bed was measured to accompany the research focused on subsurface flow

characterization. A falling head permeameter was developed for testing the vertical hydraulic conductivity through

a periphyton covered bed. The permeameter was a 4” clear acrylic tube with a ruler attached to the side. The

permeameter was placed on the bed surface and water was poured into the tube creating a very localized, elevated

piezometric head condition. The fall rate of the water column was recorded and used to determine the hydraulic

conductivity of the bed at the location of the measurement.

6.12 Conductivity Probes As part of a tracer studies conducted in the channel, nests of conductivity probes were installed in the bed. The

location of the arrays varied with the study. The probes were used along with salt injection at the upstream end of

the flume. Probe data was recorded with dataloggers.

6.13 Water chemistry/characterization Water quality sondes (Hydrolabs) were used to measure and log water quality on the upstream and downstream

ends of the test section. Measurements include water temperature, water pH, total dissolved solids, and dissolved

oxygen. They were exclusively used in the Ecogeomorphology Study. Data files used the following naming

scheme:

Hydrolab Data: Month_day_[location]_[time interval]

Where [location] was either upstream or downstream and time interval was the time between samples (default

was 1 minute if not listed). An example filename is: Jun_12_US_2min.xls. An example of a data file is shown in

Appendix V.

6.14 Subsurface water temperature Stream temperature is a critical component of aquatic habitat for salmonids and other species, and it is the

subject of regulatory scrutiny. Stream temperatures are a complex function of heat fluxes from solar radiation,

long-wave radiation, evaporative cooling, and groundwater and tributary inputs (Evans et al., 1998). While much

management concern and research focus has concentrated on land-use effects on stream temperature, recent

studies suggest that groundwater inputs and hyporheic exchange may exert significant influence on spatial and

temporal patterns of stream temperature (Poole and Berman, 2001). Field studies in small streams have shown

of 32

that the magnitudes of diurnal temperature fluctuations are correlated with the extent and residence time of

hyporheic exchange flows (Johnson, 2004; Wondzell and Johnson, 2004), which are in turn controlled by channel

morphology (Kasahara and Wondzell, 2003). To address some of the question involved in temperature

modification by hyporheic exchange, a small study of subsurface temperature was included in the

Ecogeomorphology Study of StreamLab06. Fifteen Tidbit Stowaway thermistors nested into five gravel-packed,

perforated PVC pipes were buried in the channel bed prior to the alternate bar runs of Phase 3. The thermistor

nests were placed at the heads and tails of two bars and at the cross-over channel between the bars. Bar migration

resulted in changes in relative position of the thermistor nests during and between experimental runs, and bed

scour necessitated their removal on at least one occasion. Each time the nests were reburied they were placed at

the heads, tails, and cross-overs of the existing channel morphology. Thermistors recorded data at 15 minute

intervals with ±0.2°C accuracy, and water column temperatures were recorded by the sediment flux computer and

Hydrolabs during high and low flows respectively.

6.15 Nutrient uptake Salt and nutrient pulses were injected and tracked over the various phases of StreamLab06 to determine the

effects of the geomorphology on the pathway and uptake of the material. Piezometer and surface grab samples

were taken at various locations and times following the injection to quantify the uptake. These samples were

analyzed using an autoanalyzer. Soluble Reactive Phosphorus (SRP) levels were measured at many locations

along the test reach to quantify net uptake in the flume and abiotic uptake from the sediment.

6.16 Ecological measurements A suite of water and ecological measurements were made during the Geomorphology Study. The

measurements include absorbance/chlorophyll, ash-free dry mass (AFDM),gas evasion, dissolved organic

carbon, Ammonia - NH4, and invertebrate drift.

of 32

7. DATA ARCHIVAL AND ACCESS

All the data collected in the StreamLab06 research is available to the general public. As is documented in the

preceding section, this is a vast dataset that covers four individual studies. There are many physical, biological,

and chemical datasets. Details about the format of each of the datasets are provided in the sections above. This

information is also archived along with the datafiles as meta-data. Table 7.1 summarizes the major datasets

available from StreamLab06.

Table 7.1. Summary of major dataset available in NCED data repository

The sediment flux data is perhaps the most important dataset since it contains the master time stamp for the

runs and this time was used to coordinate other data acquisition activities. Much effort has been put into reviewing

the sediment flux data to validate the robustness of the dataset. These data are presented in a raw form as well as

in various modified forms. All information about processing this data is contained in the readme.txt files and meta-

data. Appendix VI provides documentation on the post-processing procedures for the sediment flux files.

StreamLab06 is archived on the NCED Data Repository (www.repository.nced.umn.edu). The repository

allows viewing and/or download of any of the available datasets. Please contact the NCED or SAFL administrative

offices for further information on accessing data.

Dataset

Acoustic Doppler Velocitmeter Data

Backbone Data Sedflux (incl. discharge & water temp)

Biosensor and dissolved oxygen data

Conductivity data

General water chemistry data

Grainsize data

Hydrolab data

Notes and maps

Permeability data

RFID data

Surface photographs

Topographic scans

Video footage


of 32

8. CONCLUSIONS

This report provides a summary of the research projects performed under the StreamLab06 effort. The projects

were carried out between January and October of 2006 and involved a number of NCED researchers as well as

visitors from across the United States. The work utilized the newly renovated Main Channel Facility at the St.

Anthony Falls Laboratory, University of Minnesota. The overall scope of work undertaken in StreamLab06 was

large and served to test the boundaries of the facilities and the experimental conditions. The data from

StreamLab06 were collected within four well-defined research studies: 1) Ecogeomorphology Response to High

Flow Events, 2) Gravel Augmentation and Patch Dynamics Study, 3) Sand Infiltration, and 4)

Aggradation/Degradation. We briefly summarize the project scope for each of these studies however details on

the studies are not included.

This report also provides a detailed summary of the instrumentation and data collection used in the

StreamLab06 study. All datasets collected during this campaign are archived on the NCED archive at

(www.repository.nced.umn.edu). The detailed information provided in this report provides the information to

access and use this data by all interested researchers. Additionally, the facilities and instrumentation describe in

this report are permanent upgrades to the Main Channel facility and as such, the technical information provided

here will help future researchers in effectively using the facility.


of 32

APPENDIX I: References

Diplas, P., Parker, G., 1992. Deposition and removal of fines in gravel-bed streams. In: Billi, P. (Ed.), Dynamics

of gravel-bed rivers, pp. 313–329.

Elliott, A. H., and N. H. Brooks (1997), Transfer of nonsorbing solutes to a streambed with bed forms: Theory,

Water Resour. Res., 33(1), 123.

Evans, E. C., G. R. McGregor, and G. E. Petts (1998), River energy budgets with special reference to river bed

processes, Hydrological Processes, 12, 575-595.

Hondzo, M., and H. Wang (2002), Effects of turbulence on growth and metabolism of periphyton in a laboratory

flume, Water Resour. Res., 38(12), 1277, doi:10.1029/2002WR001409

Hubbell, D.W. (1987), “Bed load sampling and analysis,” Sediment transport in gravel-bed rivers, C.R. Thorne,

J.C. Bathurst, and R.D. Hey, eds., Wiley, Chichester, U.K., 89-118.

Johnson, S. L. (2004), Factors influencing stream temperatures in small streams: substrate effects and a shading

experiment, Canadian Journal of Fisheries and Aquatic Sciences, 61, 913-923.

Kasahara, T., and S. M. Wondzell (2003), Geomorphic controls on hyporheic exchange flow in mountain streams,

Water Resources Research, 39, doi: 10.1029/2002WROO1386.

Kondolf, G.M. (1997), Hungry water: effects of dams and gravel mining on river channels, Environmental

Management 21(4): 533-551.

Lisle, T.E. 1989. Using Ôresidual depthsÕ to monitor pool depths independently of discharge. Research Note

PSW-394, USDA Forest Service PSW Station, Berkeley, CA. 4pp.

Marr, J.D., Gray, J.G., (2007) StreamLab06: Ground truth testing of conventional and surrogate bedload

monitoring technologies.

Paola, C., E. Foufoula-Georgiou, W. E. Dietrich, M. Hondzo, D. Mohrig, G. Parker, M. E. Power, I. Rodriguez-

Iturbe, V. Voller, and P. Wilcock (2006), Toward a unified science of the Earth's surface: Opportunities

for synthesis among hydrology, geomorphology, geochemistry, and ecology, Water Resour. Res., 42,

W03S10, doi:10.1029/2005WR004336.

Poole, G. C., and C. H. Berman (2001), An ecological perspective on in-stream temperature: natural heat dynamics

and mechanisms of human-caused thermal degradations, Environmental Management, 27, 787-802.

Wilcock, P. R., C. H. Orr, and J. D. G. Marr (2008), The need for full-scale experiments in river science, EOS,

Transactions, Am. Geophys. Union., Vol.89, 1.

Wondzell, S. M., and S. L. Johnson (2004), Influence of hyporheic exchange flow on the temperature of a small

mountain stream, Bull. of the N. Am. Benth. Soc., 21.

K.C.Reynolds, “Notes on the Laws of Hydraulic Similitude as Applied to Experiments with Models,” Hydraulic

Laboratory Practice, p. 759, A.S.M.E, 1929.

Sheng YP (1987) On modeling three-dimensional estuarine and marine hydrodynamics. In: Nihoul JCJ, Jamart

BM (eds) Three-dimensional models of marine and estuarine dynamics. Elsevier, Amsterdam, p 35–54

Arnon, S., Packman, A.I., Peterson, C.G., and Gray, K.A., 2007, Effects of overlying velocity on periphyton

structure and denitrification, JGR-Biogeosciences, 112, G01002, DOI: 10.1029/2006JG000235.

of 32

APPENDIX II: Research team members

NCED Researchers Institution NCED Affiliation Focus

Project Manager

Jeff Marr ([email protected]) University of Minnesota Project Manger for Stream Restoration

Integrated Project, Director of Knowledge

Transfer

Faculty

Bill Dietrich University of California, Berkeley Principal Investigator Geomorphology

Jacques Finlay University of Minnesota Principal Investigator Ecology

Efi Foufoula-Georgiou University of Minnesota Principal Investigator Civil Engineering

Miki Hondzo University of Minnesota Principal Investigator Eco-Fluid Dynamics

Gary Parker University of Illinois, Urbana-

Champaign

Principal Investigator Civil Engineering/

Geomorphology

Mary Power University of California,

Berkeley

Principal Investigator Ecology

Peter Wilcock Johns Hopkins University Principal Investigator Geomorphology/ Environmental

Engineering

Post Doctoral Fellows

Nancy Brown University of Minnesota NCED Postdoc Geomorphology

Cailin Orr University of Minnesota NCED Postdoc Ecology

Jeremy Venditti University of California,

Berkeley

Post-doc Geomorphology

Graduate Students

Ben O'Connor University of Minnesota Graduate Student Eco Fluid Dynamics

Jeremiah Jazdzewski University of Minnesota Graduate Student Eco Fluid Dynamics

Peter Nelson University of California,

Berkeley

Graduate Student Geomorphology

Rebecca Stark University of Minnesota Graduate Student Ecology

Mike Limm University of California, Berkeley Graduate Student Ecology

Staff and students

Craig Hill University of Minnesota Technical Staff Geology

Sara Johnson University of Minnesota Technical Staff Geology and Civil Engineering

NCED Visitors Institution NCED Affiliation Focus

Jeff Clark Lawrence University NCED Visitor, Collaborative Investigator Geomorphology, Hyporheic

exchange

Anne Jefferson Oregon State University NCED Visitor Geomorphology, Hyporheic

exchange

Rebecca Leonardson University of California, Berkeley NCED Visitor Sand infiltration

Aleksandra Wydzga University of California, Santa

Barbara

NCED Visitor Sand infiltration

of 32

APPENDIX III: Definitions

Acoustic Doppler Velocimetry (ADV): vertical and cross-sectional profiles of the velocity/flow field and

turbulence for both plane bed and alternate bar topography.

FlowOff toposurvey: bed surface topographic scan using a Keyence laser range finder

FlowOn toposurvey: combined water surface and bed surface topographic scan using Massa sonic range finder

and submersible sonar system generating two text files of bed elevation.

script: a formatted input file, user-defined, that is read by the DAQ Carriage and contains instructions for the

movement of the carriage and instrumentation mast.

of 32

APPENDIX IV: Main Channel safety plan

of 32

APPENDIX V: Data file formats

FlowOn & FlowOff Toposcan Data (File structure is same for Keyence, Massa, or Sonar data)

NCED Data Capture - Text Output File (Comma delineated)

Massa Data [Channel 1]Taken during Sonar Scan

Massa TargetMassa Target

RunName C:\NCED\ScanData\600lps_20060530_toposcans\600lps_20060530_flowontopo_cleangravel_altbar

Date/Time 5/30/2006 11:22:39

Scan Script NameC:\NCED\SequenceScripts\FlowOn_TopoScan_5pass_10mm_CleanGravel_AltBar.scp

RunTime 0:44:39

RunSeconds2678.683

ContinuousSec2678.683

Scan Number 1

Home Encoder Offsets8657.8 103 -1608.8

Channel Transducer XYZ Offsets

Ch0 [Sonar] 300 0 1785.5

Ch1 [Massa] 100 0 1071

Ch2 [Keyence] 200 0 1071

Ch3 [Aux] 484 -12.5 1608.8

Pass Ping X Y MassaProbeZMassa TargetClockTime RunTime H:M:SRunTime SecContinuousTime Sec

1 1 9100 1369.9 282.8 951.9 11:22:37 0:44:36 2676.023 2676.023

1 2 9110 1369.9 285.5 958

1 3 9120 1369.9 285.5 956.5

1 4 9130 1369.9 285.5 962.9

1 5 9140 1369.9 285.5 966.7

1 6 9150 1369.9 285.5 960.9

1 7 9160 1369.9 285.5 957

1 8 9170 1369.9 285.5 961.3

1 9 9180 1369.9 285.5 959.1

1 10 9190 1369.9 285.5 OutOfRange

1 11 9200 1369.9 285.5 961.3

1 12 9210 1369.9 285.5 967.4

Filename & path

Date and times pass was started

X p

osition

surfa

ce e

leva

tion

y po

sitio

nPro

be e

leva

tion

Pin

g nu

mbe

r

Pas

s nu

mbe

r

Start time relative to global start time of experiment

Sensor offsets

DAQ Cart position coefficients

Script used to collect data

Data and time acquisition started

NCED Data Capture - Text Output File (Comma delineated)

Massa Data [Channel 1]Taken during Sonar Scan

Massa TargetMassa Target

RunName C:\NCED\ScanData\600lps_20060530_toposcans\600lps_20060530_flowontopo_cleangravel_altbar

Date/Time 5/30/2006 11:22:39

Scan Script NameC:\NCED\SequenceScripts\FlowOn_TopoScan_5pass_10mm_CleanGravel_AltBar.scp

RunTime 0:44:39

RunSeconds2678.683

ContinuousSec2678.683

Scan Number 1

Home Encoder Offsets8657.8 103 -1608.8

Channel Transducer XYZ Offsets

Ch0 [Sonar] 300 0 1785.5

Ch1 [Massa] 100 0 1071

Ch2 [Keyence] 200 0 1071

Ch3 [Aux] 484 -12.5 1608.8

Pass Ping X Y MassaProbeZMassa TargetClockTime RunTime H:M:SRunTime SecContinuousTime Sec

1 1 9100 1369.9 282.8 951.9 11:22:37 0:44:36 2676.023 2676.023

1 2 9110 1369.9 285.5 958

1 3 9120 1369.9 285.5 956.5

1 4 9130 1369.9 285.5 962.9

1 5 9140 1369.9 285.5 966.7

1 6 9150 1369.9 285.5 960.9

1 7 9160 1369.9 285.5 957

1 8 9170 1369.9 285.5 961.3

1 9 9180 1369.9 285.5 959.1

1 10 9190 1369.9 285.5 OutOfRange

1 11 9200 1369.9 285.5 961.3

1 12 9210 1369.9 285.5 967.4

Filename & path

Date and times pass was started

X p

osition

surfa

ce e

leva

tion

y po

sitio

nPro

be e

leva

tion

Pin

g nu

mbe

r

Pas

s nu

mbe

r

Start time relative to global start time of experiment

Sensor offsets

DAQ Cart position coefficients

Script used to collect data

Data and time acquisition started

of 32

ADV Processed file. Used WinADV software to process .ADV files (raw data)

"Processed by: WinADV32 - Version 2.024"

"Filename = 2100lps_20060512_adv_x3480_y107_z0"

"Filtering = Unfiltered"

"Traverse Options = "

"Sampling Options = All samples"

"Scaling Options = Raw data Probe (X, Y, Z)"

"WinADV Units = cm/s,cm"

"Timesecs";"Position";"Flag";"Vx_0";"Vy_0";"Vz_0";"COR0_0";"COR1_0";"COR2_0

". . . .;

0.020; ;0;65.2400;-6.2500;-

0.9000;79;74;70;61.9;59.8;64.1;209;207;214;74.3;61.9;210.

0.060; ;0;70.4200;-

11.9100;3.8600;86;88;84;58.9;58.;61.1;202;203;207;86.;59.3;204.

0.100;

;0;66.8200;7.1300;6.5100;66;88;66;58.;58.5;61.5;200;204;208;73.3;59.3;204.

0.140;

of 32

Flourometer output data

Upstream Downstream

Date Time Time Conc. [ppb] Date Time Time Flouresence

5/18/2006

13:48:56 13.816944444 127000 1.27 2006:05:18 13:49:01 13.816944444444 1.192

5/18/2006

13:49:56 13.833611111 127000 1.27 2006:05:18 13:50:01 13.833611111111 1.093

5/18/2006

13:50:56 13.850277778 127000 1.27 2006:05:18 13:51:01 13.850277777778 1.042

5/18/2006

13:51:56 13.866944444 128000 1.28 2006:05:18 13:52:01 13.866944444444 1.067

5/18/2006

13:52:56 13.883611111 126000 1.26 2006:05:18 13:53:01 13.883611111111 1.078

5/18/2006

13:53:56 13.900277778 127000 1.27 2006:05:18 13:54:01 13.900277777778 1.119

5/18/2006

13:54:56 13.916944444 130000 1.3 2006:05:18 13:55:01 13.916944444444 1.06

5/18/2006

13:55:56 13.933611111 129000 1.29 2006:05:18 13:56:01 13.933611111111 1.039

5/18/2006

13:56:56 13.950277778 133000 1.33 2006:05:18 13:57:01 13.950277777778 1.296

5/18/2006

13:57:56 13.966944444 135000 1.35 2006:05:18 13:58:01 13.966944444444 1.017

5/18/2006

13:58:56 13.983611111 131000 1.31 2006:05:18 13:59:01 13.983611111111 1.133

of 32

Hydrolab Output Datafile

Data files used the following naming scheme:

Hydrolab Data: Month_day_[location]_[time interval]

Where [location] was either Upstream or Downstream and [time interval] in time interval between samples

(default was 1 minute if not listed). An example filename is: Jun_12_US_2min.xls. An example of a datafile is

shown in Appendix V.

Date / Time Temp [°C]

pH [Units] SpCond [µS/cm]

TDS [g/l]

DO% [Sat]

DO [mg/l]

8/23/2006 12:45 25.24 9.15 389 0.2 88.8 7.1

8/23/2006 12:46 25.24 9.15 389 0.2 88.7 7.09

8/23/2006 12:47 25.27 9.15 388 0.2 84.8 6.78

8/23/2006 12:48 25.1 9.14 389 0.2 93.4 7.49

8/23/2006 12:49 25.06 9.12 389 0.2 85 6.83

8/23/2006 12:50 25.18 9.14 389 0.2 92.9 7.44

8/23/2006 12:51 25.28 9.15 389 0.2 97.1 7.76

8/23/2006 12:52 25.3 9.16 389 0.2 97.1 7.76

8/23/2006 12:53 25.3 9.16 389 0.2 96.5 7.71

8/23/2006 12:54 25.3 9.16 389 0.2 96.3 7.69

8/23/2006 12:55 25.31 9.16 389 0.2 96.1 7.67

of 32

Autoanalyzer Output Data

ID number

Date

sample

Location/type SRP

(ug/L)

NO3N

(mg/L)

323 10-Jun-06 Site 1a pre 25 3.29

324 10-Jun-06 Site 1b pre 25.5 3.32

325 10-Jun-06 Site 2a pre 29.7 3.3

326 10-Jun-06 Site 2b pre 24.3 2.72

327 10-Jun-06 Site 3a pre 29.6 3.31

328 10-Jun-06 Site 3b pre 27.9 3.37

329 10-Jun-06 Site 4a pre 28.8 3.37

330 10-Jun-06 Site 4b pre 27.8 3.37

331 10-Jun-06 Site 5a pre 24.4 3.43

332 10-Jun-06 Site 5b pre 28 3.41

333 10-Jun-06 Site 6a pre 27.3 3.45

334 10-Jun-06 Site 6b pre 28.5 3.46

335 10-Jun-06 Site 1a post 32 3.5

336 10-Jun-06 Site 1b post 34.9 3.4

337 10-Jun-06 Site 2a post 29.9 3.48

338 10-Jun-06 Site 2b post 27.1 3.53

339 10-Jun-06 Site 3a post 25.2 3.21

340 10-Jun-06 Site 3b post 26.1 2.05

341 10-Jun-06 Site 4a post 32.3 3.29

342 10-Jun-06 Site 4b post 26 3.48

343 10-Jun-06 Site 5a post 26 3.29

344 10-Jun-06 Site 5b post 27.1 3.11

345 10-Jun-06 Site 6a post 26.9 3.63

346 10-Jun-06 Site 6b post 25.2 3.27

of 32

APPENDIX VI: Sediment flux post processing

The Main Channel facility is equipped with a Sediment Flux Monitoring System (SFMS). The system is made

up of five weigh drums that hang from five individual load cells. Each load cell is monitored by the Central Data

Acquisition Computer (CDAQ). See Section 6 for a complete description of the facility. This appendix section

provides details on the post processing of the sediment flux data.

The sediment flux files are viewed as a key dataset for the StreamLab06 experiments because the data contains

the master timestamp for the runs. All other computers and flume works were coordinated off the timestamp on

this computer. For this reason and because sediment flux is a key component of the research, this data was carefully

validated and processed into various formats that we feel are usable to a variety of users. All data, including the

sediment flux data, are available on the NCED data repository (respository.nced.umn.edu).

Sediment flux Data Archival Organization The sediment flux (and other backbone data) are organized on the archive by Phase and Discharge. Each folder

contains the information in five types of forms:

a) RawData - contains the raw data from sediment flux system. This is original data collected during

run.

b) InputData by date - contains work files and the two final files used for input into MatLab. One file

exists for each day of run. The files are:

a. [discharge]lps_sedflux_[bed material]_[morphology]_final.csv

b. [discharge]lps_sedflux_[bed material]_[morphology]_final_modified.csv

c) Input data continuous - same input data as "input by date" but files are concatenated into a single .csv

file. The analysis performed by NCED and posted on archive is by date and does not use this format.

We provide this format incase others want the data this way.

d) Final - results from the analysis using matlab scripts. Many files of accumulation, flux, temp,

discharge, and associated statistics. Results are organized by day. Most runs had two days of

flooding (day1 and day2).

e) Matlab - contains the actual script files used to analyze the data. (note that separate matlab scripts are

used for each run since small variation in run parameter make this provides the greatest transparency

of our analysis methods.

Converting Raw Data into Input Data: Pre-processing steps Raw data was collected and is stored and available on the archive, but does contain normal issues associated

with raw data and needs to be "cleaned" prior to analysis. The first step in our analysis of the raw data files was

to clean and validate the data. These files are provided in the archive and have “_modified" in the filename. The

main feature of this data is that flume startup and flume shutdown periods have been removed. Flux computations

that were a part of the sediment flux acquisition were removed. Also, a continuous experimental time has been

added to the data so that consecutive days can be tracked. This experimental time is the official time for the

experiments and is used to coordinate other data acquisition and other activities in the research.

We make the modified data available in two forms. The first is by date - all startup and shutdown portions are

removed. Experimental time starts on the first day at the first moment design discharge is reached. 2nd and 3rd

days pick up the experimental time and continue it through the end of the multi-day discharge period.

The second way we provide the modified data is by discharge. These are large files that summarize all the data

collected at a discharge. In other words, for discharges lasting over several days, all data are collected in a single

file. Again, all startup and shutdown portions are removed and experimental time is added to the data.

The final steps in preparing data for input into MatLab involves removing character text in the data. "NoData"

was recorded when water surface measurements were poor quality. MatLab is not able to deal with character/data

files easily. The process for dealing with this involves manually (xls or other program) replacing "NoData" with

numeric -9999. Next, columns with data formats (mm/dd/yyyy) are removed and columns with time formats

(hh:mm:ss) are removed. A first column titled "day" is added to differentiate for continuous data changes in day.

After removing/adding these columns, the file has 11 columns of data. This file is saved with the *_modified.csv

in its name, which is required by the Matlab scripts.

of 32

MatLab Scripts Two scripts are used to process data. The first is StreamLabCumulativeWeightAlgorithm_vn.m. This script

does the following:

a. removed tipping events

b. creates a continuous accumulation dataset of weight in a drum over entire dataset within MatLab

Workspace.

c. outputs a csv file of time and water discharge

d. outputs a csv file of time and temperature

e. outputs statistical data on temp and water discharge

f. outputs a csv file of time and continuous accumulation in the five weigh drums.

g. outputs a csv file of corrected input data with -9999s replaced with values equal to preceeding WSEL

and discharge

The second script is SedfluxComp_JDMmethod_v7.m. This script does the following:

a. uses the cumulative weights computed in previous script and located in MatLab Workspace.

b. computes flux using a difference of means approach. averaging and differncing windows are 90

measurments

c. outputs csv files of computed flux for each of the five pans

d. outputs statistical file of fluxrates

Final Data The final output folder located under \sedflux files (FINAL) contains the final output from the matlab scripts

and our analysis of the sediment flux and other backbone data. The files located in this folder and a description of

what they contain are provided below.

aaaalps_yyyymmdd_Continuous_cummulative_weights.csv – comma delimited text file containing

the modified flux data with the weigh-drum tipping events removed, the start up and shut down events

removed, and other data anomalies removed. This is the input data for computing sediment flux should

someone want to compute flux on their own. The data is essentially the record of continuous

accumulation in the weigh drums over the length of the experiment.

aaaalps_yyyymmdd_fluxrates_output.csv – comma delimited text file of flux computed over the

duration of the data. Continuous_cummulative_weights.csv was used to generate this output file using

the MatLab scripts provided in the root folder of this discharge.

aaaalps_yyyymmdd_fluxstats_ouput.csv – comma delimited text file providing summary statistics of

mean flux and standard deviation at each weigh-drum for sediment flux over the duration of the dataset.

aaaalps_yyyymmdd_stats_discharge_temp.csv – comma delimited text file containing summary

statistics of means water discharge and water temperature.

aaaalps_yyyymmdd_waterdischarge.csv – comma delimited text file containing the recorded water

discharge over the duration of the dataset.

aaaalps_yyyymmdd_watertemperature.csv– comma delimited text file containing the recorded water

temperature over the duration of the dataset.

In the files listed above, aaaa is the discharge in liters per second and yyyymmdd is the date.

streamlab06: large scale ecogeomorphic flume - conservancy

Documents